DeepSeek · First MLX build 11 Dec 2024
DeepSeek-V2.5-1210
MLX
Text deepseek
A 236B parameter text generation model from DeepSeek, with 3 MLX builds, from 3-bit to 4-bit.
- Parameters
- 236B From Hugging Face
- Quantisation
- 4-bit DeepSeek-V2.5-1210-4bit
- Memory needed
- 151 GB Estimate, unified memory
- Download
- 133 GB Weights, estimated
- Downloads
- 148 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 3 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 151 GB at 4-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · mlx-lm$ pip install -U mlx-lm
$ mlx_lm.generate --model mlx-community/DeepSeek-V2.5-1210-4bit --prompt "Hello"MLX builds
3 on mlx-community| DeepSeek-V2.5-1210-3bit | 3-bit | 103 GB | 118 GB | 53 |
|---|---|---|---|---|
| DeepSeek-V2.5-1210-3bit-32 | 3-bit | 103 GB | 118 GB | 36 |
| DeepSeek-V2.5-1210-4bit | 4-bit | 133 GB | 151 GB | 148 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.